Digital Knowledge: Virginia Untold

Author: Allison Kuhnell

Over the last couple decades, LAM institutions have been soliciting the publics’ help through crowdsourcing. Some projects are limited-run exhibits which require public submissions, or new digital collections that enlist volunteers for transcriptions. ArtUK is the “online home for every public art collection”, and therefore places user involvement front and center. The National Archives has a long running “Citizen Archivist” project. For this project, I stumbled into Virginia Untold: The African American Narrative and the related public transcribing and indexing projects. 

Library of Virginia: Virginia Untold

I wanted to see if any local Virginia GLAM institutions had current crowdsourcing projects, and started by exploring the Virginia Museum of History and Culture. Virginia Unknown is linked as a partnership collection on the VMHC’s collections webpage. Virginia Untold is a digital collection of documents, mostly dating to the 18th and 19th centuries, relating to the lived experiences of enslaved and free Black and multiracial people in Virginia. Virginia Untold has digitized over 18,000 primary documents, including Free Registers, Freedom Suits, Coroner’s Inquisitions, Fiduciary Records, and Petitions. These documents are vital for researchers and educators, professional and amateur, as they contain pieces of information about thousands of individuals that were previously unseen and inaccessible. In addition to digitizing these records, the Library of Virginia is working to transcribe and index each document. Anyone can sign up and start transcribing.  

Signing up to transcribe is easy, you need to provide an email, username, and real name. Account activation is instant, there is no required training or verification. While this makes it easy for users, having such open access could make the review process more difficult for the library staff, as they may have to find and correct more mistakes. I signed up, under the username aclaire89, and began working on Arlington Country Free Registers. I currently live in Arlington so I was drawn to my county records.  

The Library of Virginia (LVA) has a number of transcription projects, all under their “Making History” initiative, and uses the crowdsourcing platform From the Page to host the actual transcription work. If you were looking at the Virginia Untold collection home page, the third display graphic on the page conveys information about getting involved as a volunteer transcriber. From the Library of Virginia homepage however, it is harder to find the transcription projects. The LVA homepage has two major “branches” for users trying to search for a book or resource; the catalog and the digital collection title “Virginia memory”. The digital collection houses digital exhibits, curated collections, and special projects. The Virginia memory site is where a user would first encounter the volunteer transcription opportunities. On the LVA webpage, links to “get involved” or volunteer only lead to donation options and more traditional volunteer roles in the Library’s physical space. 

Getting a Crowd 

It initially feels detrimental to the transcription mission to have the projects listed on a secondary site, but there are some reasons for this. The Library website has to do lots of things for lots of different users. Most people looking at the LVA online are seeking practical information about hours, books they have, or upcoming events. The digital collection is easy to find and navigate, and it makes logical sense to only display transcription projects once a user has found the digital collection and is engaging with that specific LVA content. “Making history” has been active since 2014, so the project is well established and seems to have a strong volunteer community, meaning that most projects are already mostly transcribed and new ones are attended to quickly. 

To help draw people in, the LVA characterizes volunteers as dedicated and vital. They make an effort to help people understand how important these records can be for research and other users. The project still seems geared towards people who already have some interest in local history or archives. Getting everyday users to contribute in tagging, transcription and indexing has been called “fun labor” (Agostino, 2015). Users assist the LAM institute, and in return feel satisfied and intellectually challenged and stimulated. The downside to crowdsourcing, may be the overreliance on volunteer hours and unpaid labor. LAM institutions cannot become too dependent on freely given labor, and need to still ensure projects and collections are properly and sufficiently supported, funded, and maintained. The LVA has hosted “transcribe-a-thons”, which offers a way for volunteers to feel more like they are working alongside each other and making a significant contribution, and a way for the museum to advertise transcribing. 

Indexing with LVA

Currently, LVA needs volunteers for identification, transcription, indexing, and OCR text correction. I chose to index pages in the Arlington County Free Registers. Rather than transcribing everything on the page, the user gathers any available information, such as name and age, from the digitized image of a Register page and adds it to a spreadsheet. Free registers in Virginia were written in two ways, one like a survey page with lists of names and information in columns and rows, and one in a narrative way where the clerk wrote out a paragraph about the person registering as free. I indexed several of the narrative pages. 

As you can see on the top right, once a page is saved by the volunteer, it is marked for review. LVA staff review every contribution, make edits, and mark as finalized. Staff review must take a significant amount of time, but must be less time consuming than doing it all from scratch.  You can follow different LVA users’ activity and see how quickly or often they are able to review items. There is nothing preventing anyone from signing up and misusing the project, adding in erroneous information, but a user can be removed if needed. It is not clear if the LVA has protocols in place to permanently ban a problematic user based on something like IP address, or if someone could continue to use new emails to sign up. 

I did not encounter any difficulties with indexing, once I read a few narratives and got a feel for how the clerks were recording information. Each registration follows a similar pattern of giving the person’s name, age, any identifying scars, and their status, either born free, or emancipated in some way. Before starting, I watched a video tutorial provided by the LVA on indexing the registers specifically. They also provide completed examples and written instructions about how to interpret information and copy it into the spreadsheet. I would like to try the other projects, and will probably continue to work on indexing. The one uncertainty I had was whether or not other volunteers should review or change other entries. The LVA says that users are welcome to work collaboratively and review the work of the others, before it reaches a librarian, but I did not feel confident enough in my own judgment to make changes to someone else’s entries. There is no option to say that you have “reviewed” another volunteers page, only staff can finalize and complete a page.

Looking Ahead

Working on this indexing project felt very meaningful. I have been lucky enough to have the chance to talk with people whose ancestors were enslaved, and are working to trace family history and build a genealogical tree. Prior to the 1870 census, details about African Americans were not recorded in the U.S. census, so records like the Petitions, Registers, and Suits may offer glimpses into the life of someone who is not formally recorded elsewhere. Making these records searchable and accessible is vital for genealogy and many other research projects. 

Platforms like Zoonivers and Free the Page enable LAMs of all sizes to host crowdsourcing projects. An analysis by Art Institute Chicago of Tag Along with Adler and other crowdsourcing projects shows that there are still some demographic gaps in crowdsourcing initiatives. One of the goals of many projects is to encourage wider museum participation and gather diverse voices and ideas. However, there are often “supertaggers”, repeat users who contribute disproportionately more than most users. A survey of Tag Along with Adler showed that users were more white and more educated than the greater Illonois population. Encouraging schools to get involved with crowdsourcing is one way museums can widen their volunteer pool. While some students may only see it as another requirement, other students may develop a greater interest in the work. Collaboration between LAM institutions is also key. I found Virginia Untold, an LVA collection, on the Virginia Museum of History’s website. Crowdsourcing is also a good way for museums to combine AI and human work, by having volunteers review AI or machine created transcriptions, tags, or categorization. 

Crowdsourcing for research is also important. Institutions like science museums, Departments of Natural Resources, and universities gather data collected from everyday people on wildlife, weather, language, and more. Zooniverse hosts many of the active research projects, enabling researchers to spread a much larger research net. As technology and AI advance even further, and more collections are digitized, unlocking new research potential, crowdsourcing will continue to grow and be vital to LAM institutions. Additionally, crowdsourcing for exhibits will most likely expand and increase. During the Covid-19 pandemic, many museums began actively seeking for volunteers to share stories, gather data, or record their own experiences to help capture life during a major crisis. The research possibilities through online and in person crowdsourcing are especially exciting, and will probably see rapid growth in the coming decade. 

Reference:
Agostino, Cristiano. (2015) Museum crowdsourcing as playful labor.ICOFOM Study Series, 43a, 23-37.  DOI : 10.4000/iss.545